CCR1-Mediated STAT3 Tyrosine Phosphorylation and CXCL8 Expression in THP-1 Macrophage-like Cells Involve Pertussis Toxin-Insensitive Gα_14/16 Signaling and IL-6 Release

Chemokines are a large family of low-molecular-weight cytokines that are characterized by their ability to direct the migration of leukocytes from the bloodstream to sites of inflammation (1). Leukotactin-1 (CCL15) belongs to the CC subfamily, one of four chemokine groups (CXC, CC, C, and CX₃C), as defined by their primary structures. CCL15 exerts its effect mainly via CCR1 (2, 3), which is a G protein-coupled receptor. In addition to mediating chemotaxis, CCL15 and CCR1 were shown to regulate hematopoiesis (4), angiogenesis (5), mast cell activation (6), and inflammatory diseases, including atherosclerosis (7, 8). CCR1 is expressed in myeloid progenitor cells (9) and endothelial cells (10), both of which are capable of proliferating and differentiating into mature cells. CCR1 knockout mice are defective in the trafficking and proliferation of myeloid progenitor cells (4).

Hematopoiesis and angiogenesis require transcriptional activation, which can be mediated by STAT3. STAT3 is involved in T cell proliferation induced by IL-6 (11), c-Kit–mediated stem cell factor (SCF)-independent proliferation in human leukemia cells (12), and vessel formation triggered by GM-CSF (13). Deletion of STAT3 is embryonic lethal (14). In addition, STAT3 acts as a negative regulator of inflammatory responses in hematopoietic cells. Tissue-specific deletion of STAT3 in macrophages enhances the production of inflammatory cytokines (15), whereas disruption of STAT3 during hematopoiesis leads to severe inflammatory bowel disease (16). In addition, expression of RANTES/CCL5 (a CCR1 agonist) is regulated, in part, by a transcription complex of STAT3 and NF-κB (17).

Although chemokine receptors are typically characterized as G_i-coupled receptors, there is substantial evidence to suggest that chemokines may be able to stimulate STAT3 activity through pertussis toxin (PTX)-insensitive G proteins. CCR1 agonists were found to induce gene expression of the STAT-inducible proto-oncogene, c-fos (18). c-Fos expression and transcriptional activation is induced upon Gα₁₆ activation (19), and constitutively active mutants of Gα₁₄ and Gα₁₆ were demonstrated to enhance the activity of STAT3 in cotransfection systems (20, 21). The coexistence of Gα_14/16 and CCR1, as well as their demonstrated functional coupling in THP-1 macrophage-like cells (2, 22–26), suggests that CCR1 may use Gα_14/16 to stimulate STAT3. There is increasing evidence to support a role for STAT3 activation in CCR1-mediated cellular responses. In human macrophages and macrophage-derived foam cells, CCL15 promotes the release of matrix metalloproteinase (MMP)-9 (27), which is implicated in the progression of atherosclerosis and whose expression is regulated by STAT3 (28). Moreover, CCR1 was shown to mediate IL-6 production in marrow stromal cells upon stimulation by human myeloma cells (29), whereas MCP-1/CCL2 (a CCR2 agonist) enhances IL-6 production in fibroblast-like synoviocytes from patients with rheumatoid arthritis, and the response is mediated, in part, via PTX-insensitive G proteins (30). In addition, in THP-1 macrophage-like cells, lipoprotein (31) and human placenta extracts (32) were demonstrated to promote the expression of IL-8 (CXCL8), which is capable of inducing MMP expression (33). Interestingly, both IL-6 (34) and CXCL8 (35) are STAT3-regulated cytokines, and we recently demonstrated that activation of Gα₁₆ can lead to STAT3 activation and upregulation of CXCL8 via IL-6 autocrine signaling in HEK293 and human Jurkat T cells (36). Given the importance of STAT3 in hematopoiesis and its purported involvement in inflammatory diseases, we explored whether CCR1 can indeed induce STAT3 phosphorylation and the production of IL-6 and CXCL8 through Gα_14/16-mediated signaling. Mapping of such a pathway will help to elucidate the intricate interplay between chemokines and cytokines in their regulation of complex diseases, such as atherosclerosis and rheumatoid arthritis.

The cDNAs encoding human CCR1, wild-type and constitutively active forms of Gα, and STAT3, STAT3Y705F, STAT3S727A, and p-STAT3–TA–Luc were obtained as previously described (20, 21, 25, 26, 37). HEK293 cells (ATCC CRL-1573), as well as human monocytic THP-1 (ATCC TIB-202) and U-937 cells (ATCC CRL-1593.2), were purchased from American Type Culture Collection (Rockville, MD). Human erythroleukemia (HEL) cells (ACC 11) were from the German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany). Lipofectamine Plus reagent, Alexa Fluor 488 anti-mouse IgG, Alexa Fluor 555 anti-rabbit IgG, Stealth Select RNAi for Gα₁₄ or Gα₁₆, and Stealth RNAi negative control were purchased from Invitrogen (Carlsbad, CA). PTX was obtained from List Biological Laboratories (Campbell, CA). CCL15 (68 aa), myeloid progenitor inhibitory factor-1 (CCL23) (75 aa), and human IL-6 and human CXCL8 neutralizing Abs were purchased from R&D Systems (Minneapolis, MN). Cycloheximide, STAT3 inhibitor V (Stattic), and MEK1/2 kinase inhibitor U0126 and its inactive analog U0124 were obtained from Calbiochem (San Diego, CA). Abs used in the study were products of Cell Signaling Technology (Beverly, MA).

U-937 and HEL cells were cultured in RPMI 1640 medium supplemented with 10% FBS, 50 U/ml penicillin, 50 μg/ml streptomycin, 2 mM l-glutamine, and 1 mM sodium pyruvate. THP-1 cells were cultured in the same medium supplemented with 55 μM 2-ME. THP-1 and U-937 cells were differentiated into macrophage-like cells by treatment with 20 nM PMA for 48 h. HEK293 cells were cultured in MEM supplemented with 10% FBS, 50 U/ml penicillin, and 50 μg/ml streptomycin. Primary human mast cell cultures were prepared from peripheral blood CD34⁺ progenitor using a modified protocol of Lappalainen et al. (38). Briefly, fresh peripheral blood buffy coat of healthy adult donors was provided by the Hong Kong Red Cross. Mononuclear cells were separated from peripheral blood buffy coat by density-gradient centrifugation through Ficoll-Paque Plus (GE Medical System). CD34⁺ progenitors were isolated from the mononuclear cell suspension using a MACS system (Miltenyi Biotec), according to the manufacturer’s instructions. The CD34⁺ progenitors were cultured in a medium based on IMDM containing SCF for ≥6 wk, with IL-3, IL-9, IL-6 and IL-4 added to the culture medium sequentially for different periods of time. Cells were incubated at 37°C in a humidified atmosphere of 5% CO₂ and 95% air.

HEK293 cells overexpressing CCR1 alone (CCR1/293) or expressing CCR1 together with Gα₁₄ (CCR1/Gα₁₄/293) or Gα₁₆ (CCR1/Gα₁₆/293) were generated as described previously (23, 26). Knockdown of Gα₁₄ and Gα₁₆ in THP-1 macrophage-like cells by transfection of small interfering RNA (siRNA) against Gα₁₄ and Gα₁₆ was performed as described previously (25). HEK293 cells, CCR1/Gα₁₄/293 cells, and CCR1/Gα₁₆/293 cells were seeded on six-well plates at a density of 3 × 10⁵ cells/well. Cells were transiently transfected with cDNAs encoding p-STAT3–TA–Luc, STAT3 or its phosphorylation-resistant mutants (500 ng) or wild-type or constitutively active forms of Gα (625 ng) using Lipofectamine Plus reagent.

After 24 h of transfection, cells were serum starved with 100 ng/ml PTX for 4 h and then stimulated in the absence or presence of 10 nM CCL15 for another 24 h. Subsequently, the assay medium (culture medium without FBS) was removed and replaced by 150 μl lysis buffer provided in the Luciferase Reporter Gene Assay Kit (Roche Applied Science, Penzberg, Upper Bavaria, Germany). The 6-well plate was shaken at 4°C for 30 min, and 25 μl lysate was transferred to white 96-well microplates designed for luminescent work (Nunc, Roskilde, Denmark). An additional 25 μl lysis buffer and 25 μl luciferin substrate were added to each well to initiate the reaction. Luciferase activity was determined using a microplate luminometer LB96V (EG&G Berthold, Bad Wildbad, Germany) (20, 21).

HEL cells, as well as THP-1 and U-937 macrophage-like cells, were seeded at 1 × 10⁶ cells in assay medium (culture medium containing 0.1% BSA instead of FBS) and cultured or not with 100 ng/ml PTX for 16 h. Cells were treated with cycloheximide for 2 h or with Stattic, IL-6, or CXCL8-neutralizing Ab for 30 min and then incubated in the absence or presence of chemokines at 10 nM for specific durations at 37°C. Subsequently, cells were lysed in 100 μl lysis buffer (50 mM Tris-HCl [pH 7.5], 100 mM NaCl, 5 mM EDTA, 40 mM NaP₂O₇, 1% Triton X-100, 1 mM DTT, 200 μM Na₃VO₄, 200 μM PMSF, 4 μg/ml aprotinin, and 0.6 μg/ml leupeptin) and then shaken at 4°C for 30 min. Supernatants were collected by centrifugation at 16,000 × g for 8 min. HEK293 cells were seeded on six-well plates at a density of 5 × 10⁵ cells/well, serum-starved in MEM, and lysed in 250 μl lysis buffer. Protein concentration was determined by DC protein assay kit (Bio-Rad, Hercules, CA). Eighty micrograms of proteins of each lysate was resolved by 12% SDS-PAGE and transferred to nitrocellulose membrane by electroblotting. Abs against phosphorylated STAT3 were used for the recognition of their respective phosphorylations. Fluorographs were visualized with a chemiluminescence detection kit (Amersham Pharmacia Biotech, Piscataway, NJ). Signal intensities of the immunoreactive bands were quantified using Image J software, version 1.38x (National Institutes of Health, Bethesda, MD).

CCR1/293 cells, CCR1/Gα₁₄/293 cells, or HEK293 cells transiently transfected with wild-type or constitutively active forms of Gα were collected, and whole-cell lysates were subjected to nuclear/cytosol fractionation according to the manufacturer’s protocol (Biovision, Mountain View, CA). Briefly, 1 × 10⁷ cells was resuspended in cytosol extraction buffer and centrifuged at 16,000 × g for 5 min to collect the supernatant (cytosolic fraction). The resulting pellet was resuspended in nuclear extraction buffer and centrifuged again at 16,000 × g for 10 min to collect the supernatant representing the nuclear fraction. The concentration of protein was determined using a DC protein assay kit (Bio-Rad). Eighty micrograms of proteins was separated in 12% SDS-PAGE and transferred to nitrocellulose membrane by electroblotting. Abs against caspase-3 and CREB were used as markers for cytosolic and nuclear compartments, respectively. Fluorographs were visualized with a chemiluminescence detection kit (Amersham Pharmacia Biotech).

CCR1/293 and CCR1/Gα₁₄/293 cells grown on coverslips were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, labeled with rabbit phosphorylated Tyr⁷⁰⁵ STAT3 Ab and mouse STAT3 Ab, and stained with Alexa Fluor 555 anti-rabbit IgG and Alexa Fluor 488 anti-mouse IgG. Nuclei were stained with DAPI. Confocal microscopy was performed with a Zeiss LSM 510 META, and images were deconvoluted with LSM Image Browser Rel. 4.2 (Carl Zeiss, Oberkochen, Germany).

Culture supernatants from THP-1 macrophage-like cells or HEK293 stable cell lines stimulated with CCL15 were harvested, and the presence of cytokines, including basic fibroblast growth factor (bFGF), epidermal growth factor (EGF), G-CSF, GM-CSF, IFNs (IFN-α, IFN-β, IFN-γ), IL-2, IL-4, IL-6, CXCL8, and TNF-α, were analyzed using the Procarta cytokine assay kit (Affymetrix, Santa Clara, CA). Briefly, 50 μl Ab beads was added to the prewet 96-well microtiter plate. Then, 50 μl each standard or test sample was added to the wells, followed by 25 μl detection Ab mixture. The samples were then incubated for 30 min at 25°C with gentle shaking, followed by three rounds of gentle rinsing to wash away unbound Abs. Finally, 50 μl streptavidin-PE was added and incubated for 30 min. The beads were then washed again and resuspended in 120 μl reading buffer. Analysis was performed with a Bio-Plex 200 system (Bio-Rad) using the Bio-Plex manager software (version 5.0). A minimum of 100 beads/region was analyzed. A curve fit was applied to each standard curve, according to the manufacturer’s manual, and sample concentrations were interpolated from the standard curves. For detection of CXCL8 release from human mast cells, cells cultured for ≥6 wk were counted and seeded in IMDM containing SCF overnight before sensitization with 0.5 μg/ml human myeloma IgE for 24 h. IgE-sensitized mast cells were then washed and resuspended in IMDM containing SCF. CCR1 was activated by preincubating the cells with 100 nM CCL15 for 10 min before challenging the sensitized cells with 0.25 μg/ml anti-human IgE (Sigma) for 6 h. Culture medium was harvested, and cells were removed by centrifugation at 4°C. The amount of CXCL8 released into the culture medium by human mast cells or CCL15-stimulated U-937 macrophage-like cells was measured by ELISA (BD).

The data shown in each figure represent mean ± SEM of determinations from three or more separate experiments. Statistical analyses were performed by ANOVA, followed by the Dunnett test or paired t test.

The coexistence of CCR1 (2, 25), Gα₁₄ (26), and Gα₁₆ (22, 25), as well as their functional coupling in THP-1 macrophage-like cells (23, 25), suggests that CCR1 may use Gα_14/16 to stimulate STAT3. We began the study by determining whether CCR1 can use PTX-insensitive G proteins to induce STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation in PMA-differentiated THP-1 cells. PMA-differentiated THP-1 cells have been widely used as an in vitro model of human macrophages, and lipid-laden macrophages are one of the major cell types involved in atherosclerosis. Upon PMA differentiation for 48 h, Gα₁₆, but not Gα₁₄, was upregulated by 2-fold (data not shown); the expression of CCR1 remained unaffected (data not shown), which is in agreement with the literature (39). PMA-differentiated THP-1 cells were pretreated with PTX and challenged with 10 nM CCL15 for different periods of time (15 min–24 h), and the STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation status was determined using p-STAT3–Tyr⁷⁰⁵ or p-STAT3–Ser⁷²⁷ antiserum. STAT3 Ser⁷²⁷ phosphorylation occurred rapidly (within 15 min) upon application of CCL15, and the response faded away after 6 h, whereas STAT3 Tyr⁷⁰⁵ phosphorylation started after 2 h of CCL15 treatment (Fig. 1A, left panels). The total amount of STAT3 did not change in response to CCL15 challenge (Fig. 1A), indicating that the increase in STAT3 phosphorylation at Tyr⁷⁰⁵ or Ser⁷²⁷ was not due to variations in the abundance of the protein. Similar kinetics of CCL15-induced STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation were observed in pluripotent HEL cells (Fig. 1A, right panels), which endogenously express CCR1 (40), Gα₁₄, and Gα₁₆ (20, 21). These results demonstrate that, in cells that endogenously express both CCR1 and Gα_14/16, CCL15 can stimulate STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation in PTX-insensitive and temporally distinct manners. To confirm the involvement of Gα₁₄ and Gα₁₆ in regulating CCR1-mediated signaling, THP-1 macrophage-like cells were transfected with siRNA against Gα₁₄ and Gα₁₆ and assayed for CCL15-induced STAT3 Tyr⁷⁰⁵ phosphorylation. Compared with cells transfected with control siRNA, introduction of siRNA against Gα₁₄ and Gα₁₆ into THP-1 cells reduced the level of Gα₁₄ and Gα₁₆ expression by ∼65 and ∼70%, respectively, whereas there was no detectable change in the expression of other G proteins (data not shown). After eliminating the G_i-regulated pathways by PTX treatment, the reduction in Gα₁₄ and Gα₁₆ expression by siRNA completely abrogated the CCL15-mediated STAT3 Tyr⁷⁰⁵ phosphorylation (Fig. 1B). These results suggest that the PTX-insensitive portion of the CCR1-induced STAT3 Tyr⁷⁰⁵ phosphorylation in THP-1 cells was mediated via Gα₁₄ and Gα₁₆.

To confirm the role of Gα₁₄ and Gα₁₆ in CCR1-induced STAT3 phosphorylation, we used previously established HEK293 cells (23, 26) stably expressing CCR1 alone (CCR1/293) or expressing CCR1 together with Gα₁₄ (CCR1/Gα₁₄/293) or Gα₁₆ (CCR1/Gα₁₆/293). Gα₁₄ and Gα₁₆ were prominently expressed in CCR1/Gα₁₄/293 and CCR1/Gα₁₆/293 cells, respectively, but they were absent in HEK293 cells (23, 26). The expression of CCR1 was confirmed previously by both Western blotting analysis (26) and immunofluorescence microscopy (25), whereas the coupling of CCR1 to Gα₁₄ or Gα₁₆ was verified by agonist-induced phospholipase Cβ (PLCβ) stimulation (23). Thus, these cell lines were challenged with 10 nM CCL15 for various durations. In cells expressing CCR1 alone, CCL15 stimulated STAT3 Ser⁷²⁷ phosphorylation transiently (15–45 min), whereas the STAT3 Tyr⁷⁰⁵ phosphorylation level was unaffected by CCL15 during the entire duration of the experiment (Fig. 1C), indicating that CCR1 could not stimulate STAT3 phosphorylation at Tyr⁷⁰⁵ via endogenous G proteins. The incorporation of Gα₁₄ or Gα₁₆ strengthened the STAT3 Ser⁷²⁷ phosphorylation response (Fig. 1C); the duration of the response was lengthened in CCR1/Gα₁₄/293 cells. STAT3 Tyr⁷⁰⁵ phosphorylation was also detected in cells coexpressing CCR1 and Gα₁₄ or Gα₁₆, with stimulations of ∼3.5–4-fold, but these events occurred much later, at ≥4 h (Fig. 1C).

The cell lines were then pretreated with PTX to eliminate possible G_i-mediated STAT3 phosphorylation. Application of CCL15 for 15 min weakly stimulated Ser⁷²⁷ phosphorylation of STAT3 in cells expressing CCR1 alone, which was sensitive to PTX treatment (Fig. 2A); PTX sensitivity suggests the involvement of endogenous G_i proteins in HEK293 cells. Coexpression of Gα₁₄ or Gα₁₆ with CCR1 enhanced the CCL15-induced STAT3 Ser⁷²⁷ phosphorylation, with the stimulations increased to ∼2.5-fold (Fig. 2A). Ser⁷²⁷ phosphorylation of STAT3 was resistant to PTX in CCR1/Gα₁₄/293 and CCR1/Gα₁₆/293 cells, demonstrating that CCR1 can mediate STAT3 Ser⁷²⁷ phosphorylation via Gα₁₄ and Gα₁₆ because they are the only two PTX-insensitive G proteins known to be recognized by CCR1 (25, 26). To investigate CCR1/Gα_14/16-stimulated STAT3 Tyr⁷⁰⁵ phosphorylation, the cell lines were pretreated with PTX and stimulated with CCL15 for 6 h. In cells coexpressing CCR1 and Gα₁₄ or Gα₁₆, CCL15-induced STAT3 Tyr⁷⁰⁵ phosphorylation was resistant to PTX (Fig. 2A). Another CCR1 agonist, myeloid progenitor inhibitory factor-1 (CCL23), induced STAT3 phosphorylation at Ser⁷²⁷ within 15 min, but STAT3 phosphorylation at Tyr⁷⁰⁵ was detected 6 h after drug treatment, and the response was not sensitive to PTX (Fig. 2B). These results indicated that CCR1/Gα_14/16-induced phosphorylation of STAT3 at Tyr⁷⁰⁵ and Ser⁷²⁷ have distinct kinetics. It should be noted that CCL15-induced STAT3 phosphorylation at Tyr⁷⁰⁵ was a delayed response because it remained detectable at 6 h, even when the agonist was washed out 15 min after application (data not shown).

Next, luciferase reporter assay was used to demonstrate that CCR1/Gα_14/16-mediated STAT3 phosphorylation can indeed lead to STAT3 transcriptional activity. CCR1/Gα₁₄/293 cells were transfected with cDNAs encoding STAT3-driven luciferase reporter, pretreated with PTX, and challenged with 10 nM CCL15. Consistent with the results for STAT3 Tyr⁷⁰⁵ phosphorylation, CCL15 significantly stimulated the STAT3-driven luciferase activity (Fig. 2C). Similar results were obtained in STAT3-Luc–expressing CCR1/Gα₁₆/293 cells (Fig. 2C). To investigate whether CCR1/Gα₁₄-mediated Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation of STAT3 can occur independently, CCR1/Gα₁₄/293 cells were transfected with vector plasmid, wild-type STAT3 or its phosphorylation-resistant mutants (STAT3Y705F and STAT3S727A), pretreated with PTX, and then challenged with 10 nM CCL15 for 15 min or 6 h. CCR1/Gα₁₄-induced STAT3 Tyr⁷⁰⁵ phosphorylation was attenuated in STAT3Y705F-expressing cells but not in STAT3S727A-expressing cells (Fig. 2D, left panels). In contrast, CCR1/Gα₁₄-mediated STAT3 Ser⁷²⁷ phosphorylation was abolished by STAT3S727A overexpression, whereas it was unaffected by STAT3Y705F overexpression (Fig. 2D, right panels). These results implied that CCR1/Gα₁₄ activated STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylations independently.

As a latent cytoplasmic transcription factor, STAT3 is activated by cell surface receptors and translocates from cytosol to nucleus to regulate gene expression. The ability of CCL15 to stimulate STAT3 Ser⁷²⁷ phosphorylation in the absence of Tyr⁷⁰⁵ phosphorylation is intriguing, because the Ser⁷²⁷ site is generally believed to enhance STAT3 transcriptional activity, which requires Tyr⁷⁰⁵ phosphorylation (41). Nuclear/cytosol fractionation was performed to investigate whether CCR1/Gα₁₄ activation can lead to the translocation of Tyr⁷⁰⁵- or Ser⁷²⁷-phosphorylated STAT3 to the nucleus. Consistent with the results shown in Fig. 2A, the level of STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylations could not be stimulated by CCL15 in PTX-treated CCR1/293 cells (Fig. 3A, upper right panels). Subcellular fractionation also indicated that no detectable change in STAT3 Tyr⁷⁰⁵ or Ser⁷²⁷ phosphorylation was observed in cytosolic or nuclear compartments. Coexpression of Gα₁₄ with CCR1/293 cells allowed CCL15 to stimulate STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation, even in the presence of PTX (Fig. 3A, lower right panels). Ser⁷²⁷ phosphorylation of STAT3 was increased after treatment with CCL15 for 15 min, and it gradually decreased to near basal levels by 6 h. Similar kinetics of CCL15-triggered STAT3 Ser⁷²⁷ phosphorylation were detected in the cytosolic compartment. However, CCL15-activated Ser⁷²⁷-phosphorylated STAT3 was not detectable in the nuclear compartments (Fig. 3A, lower panel). Tyr⁷⁰⁵ phosphorylation of STAT3 was elevated after treatment with CCL15 for 4 h and gradually increased to the maximal level by 6 h; a similar profile was observed in the nuclear compartment. However, CCL15-activated Tyr⁷⁰⁵-phosphorylated STAT3 was not detectable in the cytosolic compartments (Fig. 3A, lower panel). Abs against caspase-3 and CREB were used as markers for the cytosolic and nuclear compartments, respectively. These results were further confirmed by confocal microscopy to reveal the subcellular localization of STAT3 Tyr⁷⁰⁵ phosphorylation. In line with the results illustrated in Fig. 3A, CCL15 did not elicit STAT3 Tyr⁷⁰⁵ phosphorylation in CCR1/293 cells (Fig. 4). However, the introduction of Gα₁₄ to CCR1/293 cells supported the CCL15-induced STAT3 Tyr⁷⁰⁵ phosphorylation, which was primarily localized to the nucleus (Fig. 4). Thus, these results demonstrated that Tyr⁷⁰⁵-phosphorylated STAT3, but not Ser⁷²⁷-phosphorylated STAT3, was translocated to the nucleus after CCR1/Gα₁₄ activation. We demonstrated previously that members of the G_q subfamily can induce STAT3 Tyr⁷⁰⁵ phosphorylation (37). Thus, we investigated whether the activation of Gα₁₄ or Gα₁₆ could translocate Tyr⁷⁰⁵-phosphorylated STAT3 into the nucleus. In line with the previous study, active mutants of Gα₁₄ and Gα₁₆ could induce STAT3 Tyr⁷⁰⁵ phosphorylation, which was primarily localized to the nucleus (Fig. 3B).

Because CCR1/Gα_14/16-induced STAT3 Tyr⁷⁰⁵ phosphorylation occurred after prolonged treatment with CCL15 (Fig. 1), this may imply that protein synthesis of a STAT3 activator may be required. To address this possibility, HEK293 stable cell lines were pretreated with cycloheximide (a transcription/translation inhibitor) for 2 h and then incubated with CCL15 for 15 min or 6 h. Cycloheximide treatment of cells prior to CCL15 treatment led to a complete abrogation of CCR1/Gα_14/16-induced STAT3 Tyr⁷⁰⁵ phosphorylation without affecting the cell viability, whereas no inhibitory effect was observed on Ser⁷²⁷ phosphorylation (Fig. 5A). The same inhibitory effect on CCL15-induced STAT3 Tyr⁷⁰⁵ phosphorylation was observed in PMA-differentiated THP-1 cells (Fig. 5B). CCL15-activated Ser⁷²⁷ phosphorylation of STAT3 in HEK293 stable cell lines (Supplemental Fig. 1A) and THP-1 cells (Supplemental Fig. 1B) required ERK, as demonstrated by the use of an MEK1/2 kinase inhibitor (U0126) and its inactive analog (U0124). Therefore, the late and robust activation of STAT3 Tyr⁷⁰⁵ phosphorylation induced by CCR1/Gα_14/16 may require the synthesis of STAT3 activators. We then investigated whether the conditioned medium from CCL15-stimulated CCR1/Gα₁₄/293 or CCR1/Gα₁₆/293 cells was sufficient to phosphorylate STAT3 at Tyr⁷⁰⁵. Conditioned medium was collected from CCR1/Gα₁₄/293 or CCR1/Gα₁₆/293 cells stimulated with CCL15 for 15 min, 4 h, or 6 h, transferred to parental HEK293 cells, and incubated for 30 min; STAT3 Tyr⁷⁰⁵ phosphorylation was observed in the parental HEK293 cells treated with conditioned medium of CCR1/Gα₁₄/293 or CCR1/Gα₁₆/293 cells with 6 h of CCL15 treatment (Fig. 6A). Similarly, STAT3 Tyr⁷⁰⁵ phosphorylation was detected in THP-1 cells treated with conditioned medium of THP-1 cells with 6 h of CCL15 treatment (Fig. 6B). The same phenomenon was observed when the conditioned medium was collected from HEK293 cells overexpressing active mutants of G_q subfamily members (Gα₁₄, Gα₁₆, Gα_q, and Gα₁₁) but not Gα_z (the PTX-insensitive G_i subfamily members) (Fig. 6C). These results indicated that there were cytokines or growth factors in the conditioned medium that could induce STAT3 Tyr⁷⁰⁵ phosphorylation.

Many cytokines and growth factors are either STAT3 activators or their transcriptions are regulated by STAT3 activation; they include bFGF, EGF, G-CSF, GM-CSF, IFN-α, IFN-β, IFN-γ, IL-2, IL-4, IL-6, CXCL8, and TNF-α. Multiplex cytokine detection analysis was performed to screen for the aforementioned STAT3 activators or STAT3-regulated cytokines or growth factors in the conditioned medium of CCL15-stimulated CCR1/Gα₁₄/293 and CCR1/Gα₁₆/293 cells. After 4 h of drug incubation, CCL15 significantly stimulated IL-6 production in CCR1/Gα₁₄/293 and CCR1/Gα₁₆/293 cells in a PTX-insensitive manner, with maximal stimulations of 11- and 8-fold, respectively (Fig. 7A). No significant CCL15-induced IL-6 secretion was observed in CCR1/293 cells, indicating that CCR1 used Gα₁₄ and Gα₁₆ to stimulate IL-6 expression. CXCL8, a proinflammatory CXC chemokine, was also detected in the conditioned medium of CCR1/Gα₁₄/293 and CCR1/Gα₁₆/293 cells with a 4-h treatment with CCL15. CCL15 induced CXCL8 production by 22.5-fold in CCR1/Gα₁₄/293 cells, whereas the production was increased 12-fold in CCR1/Gα₁₆/293 cells (Fig. 7B). No significant CCL15-induced CXCL8 secretion was detected in CCR1/293 cells. In contrast, CCL15 treatment was unable to stimulate bFGF, EGF, G-CSF, GM-CSF, IFN-α, IFN-β, IFN-γ, IL-2, IL-4, or TNF-α expression in all three cell lines. The CCL15-stimulated THP-1 cells also produced IL-6 and CXCL8 in 2 and 6 h, respectively (Fig. 7C, 7D). Moreover, the reduction in Gα₁₄ and Gα₁₆ expression by siRNA completely abrogated CCL15-mediated CXCL8 production in THP-1 cells (Fig. 7E). This demonstrates the capability of CCL15 to express IL-6 and CXCL8 via Gα₁₄ and Gα₁₆ in a native cellular environment. We further investigated whether IL-6 expression was required for CCR1/Gα₁₄- or Gα₁₆-mediated STAT3 Tyr⁷⁰⁵ phosphorylation. CCR1/Gα₁₄/293 cells were pretreated with IL-6–neutralizing Ab for 30 min before the addition of CCL15. CCR1/Gα₁₄-induced STAT3 Tyr⁷⁰⁵ phosphorylation could be inhibited by IL-6–neutralizing Ab (Fig. 8A), showing the requirement for IL-6 in the phosphorylation of STAT3 at Tyr⁷⁰⁵. The same inhibitory effect of IL-6–neutralizing Ab on STAT3 Tyr⁷⁰⁵ phosphorylation was observed in CCR1/Gα₁₆/293 cells (Fig. 8A). However, pretreatment with CXCL8-neutralizing Ab did not affect Tyr⁷⁰⁵ phosphorylation of STAT3 in HEK293 cells overexpressing CCR1 and Gα₁₄ or CCR1 and Gα₁₆ (Fig. 8B), demonstrating that CXCL8 expression was not involved in CCR1/Gα₁₄ or Gα₁₆-mediated STAT3 Tyr⁷⁰⁵ phosphorylation. The STAT3 Tyr⁷⁰⁵ phosphorylation level was unaffected by CCL15 after the addition of IL-6– or CXCL8-neutralizing Ab to CCR1/293 cells. The requirement of IL-6, but not CXCL8, for CCL15-induced STAT3 Tyr⁷⁰⁵ phosphorylation was also illustrated in THP-1 cells (Fig. 8C). To further confirm the involvement of IL-6 in CCL15-mediated STAT3 Tyr⁷⁰⁵ phosphorylation in macrophages, U-937 macrophage-like cells were treated with IL-6–neutralizing Ab for 30 min before the addition of CCL15. PMA-differentiated U-937 cells, which endogenously express CCR1 (42) as well as Gα₁₄ and Gα₁₆ (23), are another widely used in vitro model of human macrophages (43). As shown in Fig. 8D, CCL15-induced STAT3 Tyr⁷⁰⁵ phosphorylation was inhibited by IL-6–neutralizing Ab, illustrating the requirement for IL-6 in the phosphorylation of STAT3 at Tyr⁷⁰⁵.

Because the expression of cytokines could be regulated by STAT3 activation, we investigated whether STAT3 Tyr⁷⁰⁵ phosphorylation was required for CCR1/Gα_14/16-mediated IL-6 or CXCL8 production using STAT3 inhibitor V (Stattic), which selectively inhibits the activation, dimerization, and nuclear translocation of STAT3 (44). Application of Stattic abolished CCL15-stimulated STAT3 Tyr⁷⁰⁵ phosphorylation in both CCR1/Gα₁₄/293 and CCR1/Gα₁₆/293 cells (Fig. 9A), but it did not affect Ser⁷²⁷ phosphorylation of STAT3 in these cells. The level of STAT3 phosphorylations at Tyr⁷⁰⁵ and Ser⁷²⁷ was unaffected by CCL15 after pretreatment of Stattic in CCR1/293 cells. CCL15-mediated CXCL8 secretion was abrogated by Stattic in HEK293 cells overexpressing CCR1 and Gα₁₄, but no inhibitory effect on CCL15-activated IL-6 production was observed (Fig. 9B, 9C); similar results were obtained with CCL15-treated CCR1/Gα₁₆/293 cells. No significant CCL15-induced IL-6 and CXCL8 secretion was detected in CCR1/293 cells pretreated with Stattic. For THP-1 cells, pretreatment with Stattic also diminished CCL15-induced STAT3 Tyr⁷⁰⁵ phosphorylation but not CCL15-triggered STAT3 Ser⁷²⁷ phosphorylation (Fig. 10A). The requirement of STAT3 Tyr⁷⁰⁵ phosphorylation was also demonstrated in CCL15-stimulated CXCL8 production (Fig. 10B), but not CCL15-activated IL-6 release (Fig. 10B), in THP-1 cells. In U-937 cells, application of Stattic attenuated CCL15-mediated STAT3 Tyr⁷⁰⁵ phosphorylation (data not shown) and CXCL8 release (Fig. 10C). Pretreatment with IL-6–neutralizing Ab also abolished CCL15-induced CXCL8 production (Fig. 10C), showing the involvement of STAT3 Tyr⁷⁰⁵ phosphorylation and IL-6 in CCL15-triggered CXCL8 release. The putative mechanism is depicted in Fig. 11.

Lastly, to test whether chemokine-induced CXCL8 can be observed in other hematopoietic cells, we used isolated human mast cells to examine the ability of CCL15 to stimulate the production of CXCL8. Costimulation with CCR1 enhanced anti-IgE–induced mast cell activation and degranulation (45). In the unstimulated state, human mast cells did not release CXCL8 at detectable levels. Sensitization of human mast cells by human myeloma IgE and anti-human IgE resulted in the release of CXCL8; this response was enhanced significantly in the presence of 100 nM CCL15 (Fig. 10D). Anti-IgE–induced CXCL8 production in mast cells was completely abolished in the presence of Stattic, suggesting the involvement of STAT3 (data not shown). Pretreatment with IL-6–neutralizing Ab diminished CCL15-mediated CXCL8 release (Fig. 10D), illustrating the requirement of IL-6 in CCL15-activated CXCL8 production.

Although activation of STAT3 Tyr⁷⁰⁵ phosphorylation was shown to be mediated by several chemokines, including CCL5, CCL2, and SDF-1α (CXCL12) (18, 46–48), the biochemical linkage between heterotrimeric G protein activation and STAT3 in this pathway remains poorly defined. Several reports suggest that chemokines induce receptor dimerization, resulting in the activation of the JAK–STAT pathway (reviewed in Ref. 49). However, CCR1 oligomerization has not been unequivocally established. In addition, the ability of chemokine receptors to regulate STAT3 Ser⁷²⁷ phosphorylation has not been documented. By examining the role of Gα_14/16 in transfected HEK293 cells, native HEL cells, and native THP-1 and U-937 macrophage-like cells, this study provided evidence that STAT3 Tyr⁷⁰⁵ phosphorylation is not a prerequisite for Ser⁷²⁷ phosphorylation, and IL-6 autocrine signaling is apparently involved in CCR1/Gα_14/16-mediated STAT3 Tyr⁷⁰⁵ phosphorylation, which is required for the subsequent CXCL8 production. Although CCL15 stimulated both CCR1 and CCR3 endogenously expressed in THP-1 cells (50), CCR3 is downregulated upon maturation of the monocytes to macrophages, whereas the expression of CCR1 remains unaffected (39). Thus, in the differentiated THP-1 and U-937 cells used in the current study, CCL15-mediated STAT3 phosphorylations at Tyr⁷⁰⁵ and Ser⁷²⁷ are most likely to be contributed by CCR1.

Our study clearly demonstrated that CCL15-stimulated CCR1 is capable of inducing PTX-insensitive STAT3 phosphorylations at Tyr⁷⁰⁵ and Ser⁷²⁷ only in the presence of Gα₁₄ or Gα₁₆. Moreover, the chemokine-induced phosphorylations of STAT3 at Tyr⁷⁰⁵ and Ser⁷²⁷ occurred independently (Fig. 2D) and in temporally distinct manners (Fig. 1), with Tyr⁷⁰⁵-phosphorylated STAT3 translocated to the nucleus and Ser⁷²⁷-phosphorylated STAT3 remaining in the cytosol (Fig. 3A). Apparently, CCR1-mediated STAT3 activation is a delayed response that requires the induction of IL-6 (Fig. 8A, 8C, 8D). In THP-1 cells, U-937 cells, and HEK293 cells, the biological consequence of CCR1-mediated STAT3 activation is the upregulation of yet another chemokine, CXCL8 (Figs. 9C, 10B, 10C). Fig. 11 shows a putative model of CCR1/Gα_14/16-induced STAT3 Tyr⁷⁰⁵ and Ser⁷²⁷ phosphorylation, in which the present findings are incorporated with established pathways. Upon binding of CCL15 to CCR1, Gα_14/16 becomes activated and, in turn, stimulates PLCβ, which hydrolyzes phosphatidylinositol 4,5-bisphosphate to diacylglycerol, subsequently leading to activation of the PKC/c–Src pathways. The coupling of tetratricopeptide repeat 1 to Gα_14/16 (51) activates Ras and MEK/ERK, which then induces STAT3 Ser⁷²⁷ phosphorylation (20, 21). Upon prolonged pretreatment with CCL15, activated MEK/ERK may phosphorylate transcription factors, such as Sp-1 (52), leading to the translocation of transcription factors to the nucleus, upregulation of IL-6, and transactivation of IL-6R. The secreted IL-6 induces STAT3 phosphorylation at Tyr⁷⁰⁵ in an autocrine fashion, resulting in dimerization and nuclear translocation of Tyr⁷⁰⁵-phosphorylated STAT3 and subsequent CXCL8 transcription.

We showed previously that activation of Gα_14/16 by a variety of G protein-coupled receptors, including the δ-opioid, C5a, formyl peptide, and opioid receptor-like receptors (20, 21, 53), can lead to STAT3 Tyr⁷⁰⁵ phosphorylation within 15 min in HEK293 cells. In contrast, CCR1/Gα_14/16-mediated Tyr⁷⁰⁵ phosphorylation of STAT3 was only detected after prolonged drug pretreatment. This discrepancy is puzzling, especially because the receptors were expressed in the same cellular background. Although we do not have a plausible explanation, it should be noted that delayed STAT3 Tyr⁷⁰⁵ phosphorylation is not unique to CCR1; it has been similarly observed with G₁₆-coupled melatonin MT₁ and MT₂ receptors (36). Moreover, distinct temporal patterns of STAT3 phosphorylations at Tyr⁷⁰⁵ and Ser⁷²⁷ have been documented. In murine macrophage-like RAW 264.7 cells, STAT3 Ser⁷²⁷ phosphorylation induced by LPS can be observed at 5 min, whereas STAT3 Tyr⁷⁰⁵ phosphorylation requires 2 h of treatment (54). Likewise, two groups demonstrated that angiotensin II induces delayed STAT3 Tyr⁷⁰⁵ phosphorylation in rat aortic smooth muscle cells (55) and rat cardiomyocytes (56). Fukuda and coworkers (57) also showed that the delayed STAT3 Tyr⁷⁰⁵ phosphorylation stimulated by angiotensin II is mediated by autocrine/paracrine secretion of IL-6.

The induction of IL-6 and CXCL8 by CCR1/Gα_14/16 signaling has several implications. IL-6 is a potent inflammatory cytokine that directly activates STAT1 and STAT3. The production of IL-6 in CCL15-stimulated THP-1 cells and transfected HEK293 cells implied that CCR1/Gα_14/16 could also stimulate STAT1, which is in line with our previous study showing that the constitutively active mutant of Gα₁₆ can enhance the activity of STAT1 (19). Our demonstration of CCR1/Gα_14/16-mediated STAT3 Tyr⁷⁰⁵ phosphorylation leading to subsequent CXCL8 production is in agreement with the report that CXCL8 expression is transcriptionally upregulated by STAT3 in human melanoma cells (58). In addition, it was reported that JAK/STAT3 is involved in thrombin-induced CXCL8 secretion in human dermal fibroblasts (59). CXCL8 was reported to activate MMP expression (33) and suppress tissue inhibitor of metalloproteinases expression (60), and these regulations by CXCL8 are responsible for the rupture of atherosclerotic plaques. In human macrophages and macrophage-derived foam cells, CCL15 is capable of inducing the production of MMP-9 (27), whose expression is regulated by STAT3 (28), whereas CXCL8 is apparently involved in the recruitment of human NK cells to sites of early viral infection by mast cells (61) and metastasis of colon cancer (62). The release of CXCL8 activates CXCR2, which is coexpressed with CCR1 in macrophages (39) and human mast cells (63, 64), leading to STAT3 Tyr⁷⁰⁵ phosphorylation (65). Hence, the production of CXCL8 mediated by CCR1/Gα_14/16 may create another loop of STAT3 activation, which synergizes with the effect on CXCL8 secretion.

Although the Ser⁷²⁷ site is generally believed to enhance the transcriptional activity of Tyr⁷⁰⁵-phosphorylated STAT3 (41), recent studies demonstrated that STAT3 phosphorylation at Ser⁷²⁷ alone can trigger a variety of biological activities. STAT3 Ser⁷²⁷ phosphorylation is essential for postnatal survival and growth, as well as thymocyte proliferation, as shown in knock-in mice with alanine substituted for Ser⁷²⁷ in STAT3 (66). A recent study in mouse pro-B cells showed that mitochondrial STAT3 Ser⁷²⁷ phosphorylation is important in regulating the oxidative phosphorylation of mitochondria via association with complex I/II of the electron transport chain (67). We conducted mitochondria/cytosol fractionation to examine whether CCR1/Gα₁₄ activation can modulate STAT3 Ser⁷²⁷ phosphorylation in the mitochondria; our preliminary data showed that the basal level of mitochondrial Ser⁷²⁷-phosphorylated STAT3 in CCR1/Gα₁₄/293 cells was increased compared with cells expressing CCR1 alone. This tends to suggest that the basal cellular ATP level may be higher in CCR1/Gα₁₄/293 cells. The elevated basal level of mitochondrial Ser⁷²⁷-phosphorylated STAT3 was apparently unaffected by CCL15 treatment. Because there are no data on chemokines regulating cellular metabolism, further investigation is warranted to examine the biological function of CCR1/Gα_14/16-mediated STAT3 Ser⁷²⁷ phosphorylation.

In summary, IL-6 autocrine signaling is apparently involved in CCR1/Gα_14/16-mediated STAT3 Tyr⁷⁰⁵ phosphorylation, which is required for the subsequent production of CXCL8. In turn, CXCL8 may activate CXCR2, leading to further STAT3 Tyr⁷⁰⁵ phosphorylation, which synergizes with the effect of CCR1 agonists. To our knowledge, this study represents the first demonstration of chemokine-stimulated CXCL8 production via CCR1/Gα_14/16/STAT3-signaling pathways. It remains to be determined whether autocrine signaling of IL-6/STAT3/CXCL8 by CCR1/Gα_14/16 plays a role in the progression of diseases, such as atherosclerosis.

We thank Dr. Rodger A. Allen for kindly providing the human CCR1 cDNA.

The authors have no financial conflicts of interest.